Yeast chromatin assembly complex 1 protein excludes nonacetylatable forms of histone H4 from chromatin and the nucleus

Abstract

In yeast, the establishment and maintenance of a transcriptionally silent chromatin state are dependent upon the acetylation state of the N terminus of histone proteins. Histone H4 proteins that contain mutations in N-terminal lysines disrupt heterochromatin and result in yeast that cannot mate. Introduction of a wild-type copy of histone H4 restores mating, despite the presence of the mutant protein, suggesting that mutant H4 protein is either excluded from, or tolerated in, chromatin. To understand how the cell differentiates wild-type histone and mutant histone in which the four N-terminal lysines were replaced with alanine (H4-4A), we analyzed silencing, growth phenotypes, and the histone composition of chromatin in yeast strains coexpressing equal amounts of wild-type and mutant H4 proteins (histone H4 heterozygote). We found that histone H4 heterozygotes have defects in heterochromatin silencing and growth, implying that mutations in H4 are not completely recessive. Nuclear preparations from histone H4 heterozygotes contained less mutant H4 than wild-type H4, consistent with the idea that cells exclude some of the mutant histone. Surprisingly, the N-terminal nuclear localization signal of H4-4A fused to green fluorescent protein was defective in nuclear localization, while a mutant in which the four lysines were replaced with arginine (H4-4R) appeared to have normal nuclear import, implying a role for the charged state of the acetylatable lysines in the nuclear import of histones. The biased partial exclusion of H4-4A was dependent upon Cac1p, the largest subunit of yeast chromatin assembly factor 1 (CAF-1), as well as upon the karyopherin Kap123p, but was independent of Cac2p, another CAF-1 component, and other chromatin assembly proteins (Hir3p, Nap1p, and Asf1p). We conclude that N-terminal lysines of histone H4 are important for efficient histone nuclear import. In addition, our data support a model whereby Cac1p and Kap123 cooperate to ensure that only appropriately acetylated histone H4 proteins are imported into the nucleus.

FIG. 1.

Histone H4 variant proteins can be resolved on polyacrylamide gels. (A) Coomassie-stained gel showing the different mobilities of the wild-type and mutant histone H4 proteins from the H4-WT (YJB344), H4-4A (YJB347), H4-4Q (YJB2861), and H4-4R (YJB348) strains. Nuclear extracts were prepared from exponentially growing strains and separated by reverse-phase HPLC, and fractions containing histones H2A and H4 were separated by SDS-18% PAGE. Histone H2A coelutes with H4 during HPLC and serves as a loading control (64). Lanes 2 and 3 contain mixtures of extracts from the H4-WT and H4-4A strains. Lane 2 contains approximately 75% H4-WT and 25% H4-4A, and lane 3 contains approximately 75% H4-4A and 25% H4-WT as determined by densitometry of the stained gel. (B) Western blot assay of WCEs of the same strains as in panel A separated by SDS-22% PAGE and probed with the anti-H4-17-33 antibody.

FIG. 2.

(A) Coomassie-stained gel of yeast strains expressing H4-WT and H4-4A histone proteins from centromere plasmids in the H4-WT (YJB344), H4-4A (YJB347), and H4-4A/WT (YJB2318) strains. Nuclear extracts of exponentially growing strains were prepared and separated by reverse-phase HPLC. Fractions containing histones H2A and H4 were separated by SDS-22% PAGE. (B) Diagram representing the different histone H4 mutant strains constructed as described in Materials and Methods. HHF1, HHF2, and hhf2-4A code for histone H4 proteins; HHT1 and HHT2 encode histone H3; and CaURA3 encodes the C. albicans URA3 gene that was used to disrupt the chromosome XVI histone gene locus.

FIG. 3.

Strains expressing both H4-WT and H4-4A histone proteins display defective telomeric silencing. Liquid cultures of exponentially growing yeast strains with the indicated genotypes were serially diluted, spotted onto SDC−Ura and 5-FOA plates, and grown for 3 days at 30°C. Strain YJB8002 was used for the H4-WT strain, YJB8003 was used for the cac1Δ H4-WT strain, and YJB8004 was used for the H4-4A/WT strain. Wild-type strains form colonies on both SDC-Ura and 5-FOA plates because of TPE on URA3 expression.

FIG. 4.

The ratio of H4-WT to H4-4A differs in WCEs and nuclear extracts. (A) WCEs (lysates) were prepared from exponentially growing H4-WT (YJB7349), H4-4A (YJB7354), and H4-4A/WT (YJB7348) strains, and extracted nuclear material (pellet) was prepared from the H4-4A/WT strain. Both WCEs and nuclear extracts were separated by SDS-22% PAGE and probed with the anti-H4-17-33 antibody. Note that the H4-4A/WT strain (lane 4) does not contain as much H4-4A protein in the nuclear pellet as in the lysate (lane 3). Quantitation of the amounts of histone H4 present in the nuclear pellet compared to the lysate indicated that 40% of the expected H4-4A is absent in the nuclear pellet. (B) Western blot assay of nucleus-enriched preparations. Lysate, pellet (nuclei), and supernatant (Sup; cytoplasm) material from the nuclear extract protocol were separated by SDS-8% PAGE and probed with either an anti-PGK1 antibody or an anti-RAP1 antibody.

FIG. 5.

H4-4A-GFP protein is excluded from nuclei. Panels: A, H4-WT-GFP (YJB8342); B, H4-4A-GFP (YJB8343); C, H4-4R-GFP (YJB9012); D, H4-K5,12R-GFP (YJB9011); E, H4-K8R-GFP (YJB9010). Yeast strains containing the indicated GFP fusion proteins were grown overnight at 30°C on solid medium to repress expression of the GFP fusion protein. Cells were then introduced into liquid medium to induce expression of the GFP fusion protein and grown for 6 h at 30°C. Differential interference contrast (left panels) and fluorescence (right panels) micrographs of unfixed cells are shown.

FIG. 6.

Cac1p is responsible for the exclusion of H4-4A histone from chromatin. Nucleus-enriched preparations of exponentially growing wild-type (YJB7348), kap123Δ (YJB8405), cac1Δ (YJB8398), cac2Δ (YJB8817), asf1Δ (YJB8396), hir3Δ (YJB8400), and nap1Δ (YJB8402) cells were separated by SDS-22% PAGE. Western analysis was performed with the anti-H4-17-33 antibody. Densitometry was performed on the indicated number (n) of isolates, and the average amounts of H4-4A excluded from nuclear extracts were calculated. Note the increased amount of H4-4A histone protein in the cac1Δ strain.

FIG. 7.

H4-4A-GFP protein is excluded from nuclei in cac1 mutant cells. (A) H4-WT-GFP expressed in wild-type (YJB8342), cac1Δ (YJB8432), cac1Δ cac2Δ (YJB9013), and cac1Δ cac3Δ (YJB9015) cells. (B) H4-4A-GFP expressed in wild-type (YJB8343), cac1Δ (YJB8378), cac1Δ cac2Δ (YJB9014), and cac1Δ cac3Δ (YJB9016) cells. Yeast strains containing the indicated GFP fusion proteins were grown overnight at 30°C on solid medium to repress expression of the GFP protein. Cells were then introduced into liquid media to induce expression of the GFP fusion protein and grown for 6 h at 30°C. Differential interference contrast (left panels) and fluorescence (right panels) micrographs of unfixed cells are shown. Loss of the largest subunit of CAF-1 results in a subtle mislocalization of H4-WT-GFP, while loss of multiple CAF-1 subunits does not further alter the distribution of either H4-WT or H4-4A-GFP.

FIG. 8.

Overall chromatin organization is similar in cells expressing H4-WT and H4-4A histone H4. (A) ChIP was performed as described in Materials and Methods. Fifteen microliters of each PCR product was analyzed on 1% agarose gels. Lanes: 1, 1-kb DNA ladder; 2, PCR of genomic DNA; 3 to 6, YJB7650 (H4-4A); 7 to 10, YJB7651 (H4-WT);11 to 14, YJB7902 (H4-4A/WT); 3, 7, and 11, PCRs of total immunoprecipitate; 4, 8, and 12, no-antibody control; 5, 9, and 13, immunoprecipitation with anti-acetyl-histone H3; 6, 10, and 14, immunoprecipitation with anti-HA antibody. (B) Protein ChIP. Formaldehyde cross-linked lysates were isolated from the H4-4A/WT strain containing SIR3-HA (YJB8651). Active chromatin was immunoprecipitated with the anti-acetyl-histone H3 antibody; silent chromatin was precipitated with an anti-HA antibody against Sir3-HA. Nuclear extracts (lysates) and immunoprecipitates were separated by SDS-22% PAGE, and Western analysis of histone H4 proteins was performed with anti-H4-17-33 antibody. To detect the histones associated with silent chromatin, the blot, indicated by an asterisk, was exposed approximately 10 times longer than for active chromatin.